Dietary factors and coronary heart disease*

Bull. Org. mond. Sa)ut 1970, 42, 103-114 Buill. Wid Hith Org. Dietary Factors and Coronary Heart Disease* R. MASIRONI 1 Mortality data from arteriosclerotic and degenerative heart disease (AHD) and per capita consumption oftotalfat, saturatedfat, sucrose, simple sugars, complex carbohydrates, and protein, and calorie intake for 37 countries were statistically evaluated to investigate possible relationships between dietary factors and incidence of AHD. On a geographieal basis, consumption of total and saturated fats is strongly and positively correlated with the death rates, while consumption of complex carbohydrates is negatively correlated. No correlations were found with temporal changes in death rates or with differences within one country. These findings are discussed in the light of the works of many other investi- gators. It is concluded that the relation of diet to AHD is still controversial, and that the development and severity of the disease canntot be confidently attributed to any fingle dietary factor nor to blood cholesterol. The contributing effects of other factors, sueh as physical activity, mental stress, and affluence, are also discussed. It is common knowledge that mortality rates from arteriosclerotic and degenerative heart disease (AHD) are much higher in the highly industrialized, technically advanced countries than in the develop- ing countries. It is also well known that death rates from this disease have increased over the past several years, particularly in the younger age-groups: in some countries they have increased by as much as 2 or 3 times in one decade. Part of the increase is likely to be an artefact brought about by the changing fashion in certifying heart disease as a cause of death, by better diagnostic procedures, and by the longer life expectancy; nevertheless, several authors (Lew, 1957; Oliver & Stuart-Harris, 1965; Reader & Wynn, 1966) have pointed out that much of this increase cannot be accounted for simply in terms of artefacts and must be considered as real. Recent autopsy studies have confirmed that differences in aortic and coro- nary atherosclerosis between different countries and population groups are real and striking (Tejada et al., 1968). In an attempt to explain these differences much epidemiological, clinical and experimental work has been carried out by numerous investigators to deter- mine whether dietary factors play an important role in the development of AHD. * This article will also be published, in Spanish, in the Boletin de la Oficina sanitaria panamericana. 1 Scientist, Cardiovascular Diseases, World Health Organization, Geneva, Switzerland. The aim of the present article is to supply up-to- date information on AHD death rates as related, on a geographical basis, to the intake of various dietary factors. Although this kind of statistical and epide- miological evidence does not permit definitive con- clusions to be drawn as to the causal significane of the correlations detected, it is useful in identifying trends and in suggesting specific problems worthy of further study. METHODS Death rates per 100 000 population for ast*rio- sclerotic and degenerative heart disease were taken from World health statistics annuals (World Health Organization, 1958, 1968). Various groups, defined by age and sex, were studied: all ages and both sexes, 45-54-year-old males, 55-64-year-old males, and 55-64-year-old females. Dietary data were taken from the Food balance sheets published by the Food and Agriculture Organization (1949, 1966). When some data, such as those relating to saturated fats or to complex carbohydrates and simple sugars, were not directly available, they were calculated on the basis of the Food composition tables published by the Food and Agriculture Organization (1953) following the defi- nitions of Jolliffe (1961) and of Lopez et al. (1966). Although the data in the FAO Food balance sheets are a crude estimate of the food consumption, they proved to be similar to other data obtained in 2458 - 103 - R. MASIRONI accurate dietary surveys, recently carried out by a team of investigators in Finland, Greece, Italy, the Netherlands, and Yugoslavia (Den Hartog et al., 1968). The national per capita income and per capita energy consumption data referred to in Table 1 were taken from the United Nations Statistical yearbook (1967) and from the United Nations Year- book of national account statistics (1966). RESULTS Table 1 shows the death rates from AHD and the dietary consumption data for 37 countries. The statistical evaluation shows that consumption of total fat and of saturated fat (as a percentage of the total calorie intake) is strongly and positively correlated with death rates, whereas consumption of complex carbohydrates is negatively correlated. The intake of calories, sucrose, and simple sugars is correlated less strongly, and protein consumption shows a poor correlation. The values relating to the previous decade, shown in Table 2, also show similar types of correlations: for this period data were available for only 25 countries. These trends are present in all the age-groups studied; in the females 55-64 years of age, however, some of the correlations are weaker, thus suggesting the possible involvement of a sex factor. The influence of sex differences on cardiovascular mortality patterns is extensively described in the monograph by de Haas (1964). Table 3 shows the correlation coefficients between the various dietary factors listed in Table 1. The highest correlation is between total fats and saturated fats. Contrary to Yudkin's finding (1964), the present data show only a weak correlation between consumption of fats (both total and saturated) and sugars (sucrose and simple sugars). Consumption of complex carbohydrates showed a strong negative correlation with the 4 dietary constituents men- tioned above: the strong negative correlation with fat intake indicates that the negative correlation, shown in Table 1, between complex carbohydrate consumption and AHD death rates may actually be secondary to the positive correlation between fat consumption and death rates. In other words, fat may be the dietary constituent playing an important role in cardiovascular mortality while the complex carbohydrates may be just neutral bystanders. The present results are in line with the generally accepted hypothesis that excessive intake of fats, especially of the saturated type, may have a detrimen- tal effect on cardiovascular health and is therefore associated with higher cardiovascular mortality rates. The results agree also with those reported by Lopez et al. (1966) with the exception of those concerning simple sugars: whereas the present investigation did not reveal any strong correlation between the con- sumption of simple carbohydrates and cardiovascular mortality rates, Lopez et al. found a high correlation coefficient. The reason for this discrepancy may be that these authors used data relating to only 14 Euro- pean countries, in which simple sugar consumption is almost the same, i.e., 13.5 %-21.2% of total calorie intake, whereas in the 37 countries considered in the present investigation simple sugar consumption ranges from 5.1 % to 38.5% of the total calorie intake. As there are countries, such as Colombia, Costa Rica, Jordan, Mexico and others, where death rates from AHD are low and yet the calorie contribution from simple sugars is as high as, or even higher than in the European countries, then we may assume that the relation between AHD death rates and simple sugar intake, described on a geographical basis by some authors (Lopez et al., 1966; Yudkin, 1964) is actually a secondary association for which a causal significance has not been proved yet. Very good positive correlations are also present between AHD death rates and national per capita income and energy consumption, as well as between income and consumption of total fat, and saturated fat: there is also a negative correlation between income and consumption of complex carbohydrates. The implications of these findings will be discussed in more detail later. The next step in the present study was to determine whether the increase in cardiovascular mortality rates which has occurred in practically all countries over the past several years has been accompanied by a parallel increase in dietary intake. Only for 23 countries was it possible to collect both mortality and dietary data for the past 15-20 years. Between 1955 and 1965 the death rates from AHD increased on average by 51 % (all ages, both sexes); consump- tion of total fat, saturated fat, sucrose, simple sugar and, to a lesser extent, calorie intake increased also. Complex carbohydrate consumption decreased, and protein consumption decreased slightly (Table 4). This is a well established picture that has been observed by many investigators and it is considered to be a reflection of increased prosperity. The statistical analysis, however, reveals that the correla- tions between the changes observed in dietary intakes 104 TABLE 1 GEOGRAPHICAL CORRELATION BETWEEN DEATH RATES FROM AHD, DIETARY FACTORS, AND FACTORS ASSOCIATED WITH LEVEL OF NATIONAL DEVELOPMENT, 1960s Dietary factors AHD death rates a (approximate percentage of total Total calorie intake b) Country calorie In- Ener- 55-64 55-64 45-54 sBeoxesh Fat tusat Sucr- Simple Compe ro-Males, Females, Males, ~~Fat turated r Sipe carbo- Pr oe yyears years years !alxaes, ft oe sugars hyrtsteinof age of age of age a ages fat drates Jordan 49 I 19 15 6 2 230 18.5 3.7 11.0 18.6 57.9 11.1 192 300 El Salvador 51 28 17 6 2 030 19.5 5.5 - 13.8 55.0 11.2 236 183 Taiwan 58 53 16 10 2 350 15.3 4.8 4.2 5.1 71.6 9.9 185 706 Philippines 89 40 38 14 1 840 9.6 3.7 6.6 9.6 65.9 13.7 221 207 Guatemala 115 84 30 18 2 080 15.1 4.1 - 13.5 61.5 10.6 281 213 Mexico 122 65 45 18 2 610 24.8 7.9 - 17.3 45.7 11.0 412 997 Greece 162 61 57 88 2 940 26.5 5.4 4.9 12.7 55.1 13.0 570 831 Japan 165 92 51 58 2 230 14.5 3.4 7.3 8.6 70.8 12.4 694 1 954 Panama 170 76 60 36 2 310 22.6 6.8 8.7 19.6 52.2 10.0 425 1143 Spain 171 71 56 68 2800 27.9 7.5 7.0 12.7 56.4 10.9 594 1104 France 206 59 74 82 3 050 37.0 14.5 10.4 13.9 44.8 13.0 1 448 3 019 Colombia 218 130 68 31 2160 21.6 8.1 8.0 24.2 53.1 9.6 237 534 Costa Rica 224 150 87 44 2430 18.6 6.4 13.1 38.5 31.8 8.9 352 312 Romania 255 152 75 141 3160 21.3 4.4 4.3 8.0 63.7 12.3 353 2 072 Portugal 259 114 64 116 2 550 23.7 6.2 8.0 13.0 62.1 11.5 353 532 Yugoslavia 267 184 71 121 3 030 22.3 3.2 5.2 8.6 63.2 12.4 720 1 202 Poland 281 114 99 83 3 350 27.2 4.6 - 11.1 60.1 11.1 550 3 608 Chile 321 184 106 68 2 410 19.4 7.5 - 16.6 57.0 12.7 515 1113 Mauritius 322 98 178 47 2 330 18.9 3.6 17.6 15.1 50.6 8.1 213 171 Venezuela 342 230 139 49 2 310 22.6 6.1 15.6 19.0 52.2 10.3 745 2 469 Italy 387 152 134 205 2 730 26.4 7.7 8.9 14.5 53.6 11.8 884 1 961 Switzerland 395 68 124 229 3 220 36.4 13.4 14.0 21.2 35.9 11.2 1 929 2 698 Belgium 419 124 172 149 3 040 41.1 12.8 11.0 13.5 42.8 11.2 1428 4 636 Hungary 453 234 157 254 3 030 29.4 13.7 - 13.2 52.6 12.1 740 2 825 Austria 478 150 153 244 2 970 33.7 9.7 13.0 19.9 42.9 11.7 970 2 692 Uruguay 485 173 159 167 3 220 36.6 16.4 - 16.7 35.9 12.9 537 843 Sweden 490 152 130 313 2 990 40.6 14.5 14.6 20.1 33.2 11.2 2 000 5 080 Netherlands 503 136 180 188 2 950 39.5 11.5 15.9 20.7 38.6 10.6 1 274 3 508 Federal Republic of Germany 549 175 194 220 2 960 38.8 14.6 10.8 17.6 42.1 10.9 1 455 4 267 Norway 583 138 182 248 2 930 39.2 12.4 16.7 19.1 36.3 11.0 1451 3 963 Denmark 586 172 176 290 3 370 41.5 16.1 15.5 18.6 35.2 11.0 1 677 4 515 Israel 626 306 201 169 2 820 28.9 6.9 12.0 20.6 43.2 11.9 1076 2 248 United Kingdom 743 208 263 322 3 280 39.2 15.1 15.9 18.8 35.1 10.9 1467 5139 Canada 832 253 306 247 3 020 40.8 16.3 15.5 17.1 29.8 12.0 1 824 7 878 New Zealand 889 309 288 251 3 490 40.4 19.2 12.6 18.2 29.7 12.0 1 695 2 644 USA 933 301 357 316 3100 41.5 14.3 14.5 20.5 27.3 11.7 2 910 9 595 Australia 942 303 327 339 3140 38.0 16.3 14.0 21.1 30.3 11.5 1 640 4931 Finland 1 037 252 450 272 3110 33.0 16.2 13.6 18.6 42.7 12.1 1 393 2 838 Correlation coefficients of death rate to dietary factors, income and energy consumption Males, 55-64 years of age 0.69 0.74 0.80 0.66 0.31 -0.74 0.11 0.75 0.72 Females, 55-64 years of age 0.60 0.55 0.62 0.64 0.33 -0.59 0.11 - - Males, 45-54 years of age 0.59 0.70 0.74 0.64 0.32 -0.71 0.08 0.73 0.71 Both sexes, all ages 0.77 0.84 0.81 0.56 0.24 -0.72 0.21 0.83 0.77 Correlation coefficients of income to dietary factors and energy consumption 0.67 0.86 0.79 0.58 0.28 -0.76 0.18 0.89 a Death rates for 1965 per 100 000 population. b Approximate daily per capita consumption, 1960-62. Figures collated from various sources. c National per capita income in US $, 1965. d National per capita consumption in kg of coal equivalent, 1966. 106 R. MASIRONI TABLE 2 GEOGRAPHICAL CORRELATIONS BETWEEN DEATH RATES FROM AHD AND DIETARY FACTORS, 1940s-1950s AHD death rates a2 Dietary factors(approximate percentage of total calorie intake b) Countries ~~~~~~~~TotalICountries Males Males Both calorie Unsatu- complexBoth4554intake Simple Poe years years sexes, Fat irated Sucrose sgr cro rtiall ages fat suas hydrates of age of age Yugoslavia Colombia G reece Poland France Japan Ceylon c Chile Italy Hungary Uruguay Belgium Norway Netherlands Switzerland Iceland d Federal Republic of Germany Den mark Sweden United Kingdom New Zealand A ustralia Canada Finland USA 92 94 101 131 144 145 159 292 317 328 343 344 355 357 377 393 20 39 20 56 49 58 63 77 97 109 135 132 113 115 117 148 413 134 423 129 430 107 597 191 712 224 788 247 790 277 847 305 896 342 33 2144 16 1 990 66 2 266 21 2 363 54 2 357 43 1 670 29 1 680 46 2 352 157 2 249 147 2 432 99 2 478 134 2 667 153 2 899 158 2 693 235 3 106 124 3096 175 2 189 201 3 125 234 2 871 313 2 968 229 3 214 244 3 262 228 3 161 201 2 617 287 3 244 14.3 4.3 18.1 7.1 25.2 5.3 14.6 5.9 25.6 11.3 5.8 1.1 15.5 3.1 15.2 6.6 21.3 5.7 17.1 6.4 30.5 13.8 29.6 12.3 34.1 10.6 28.6 12.2 27.6 11.3 37.2 15.0 13.7 5.5 35.7 14.4 34.6 15.7 31.3 12.1 38.0 18.2 34.3 15.4 38.0 17.3 24.6 11.3 38.0 15.4 1.8 5.2 7.1 28.6 4.0 6.7 - 3.0 6.9 8.6 2.1 2.0 9.0 8.3 13.2 3.1 6.0 _ 7.0 13.5 10.3 12.1 8.8 12.1 10.2 12.6 12.2 22.1 12.9 12.1 8.0 12.0 11.7 14.2 14.8 19.0 13.7 16.0 14.9 18.0 14.3 20.3 13.8 21.2 6.1 11.0 15.0 21.0 71.1 47.1 58.7 87.0 68.6 85.0 68.1 62.1 66.1 70.8 39.4 58.6 50.9 59.6 48.7 40.9 74.7 48.1 43.1 48.3 32.1 35.1 35.3 59.7 32.0 12.1 10.2 12.0 11.2 12.9 12.0 11.9 12.4 12.6 12.0 14.5 11.8 12.5 11.5 12.0 14.7 13.7 13.8 15.7 12.0 12.5 12.9 12.0 13.9 12.2 Males, 55-64 years of age Males, 45-54 years of age Both sexes, all ages Correlation coefficients of death rate to dietary factors 0.75 0.63 0.73 0.65 0.45 -0.62 0.26 0.70 0.62 0.69 0.62 0.47 -0.62 0.23 0.81 0.70 0.70 0.75 0.52 -0.63 0.31 a Death rates for 1955 per 100 000 population. b Approximate daily per capita consumption, 1947-49. Figures collated from various sources. c Not included in Table 1 due to lack of mortality data for 1965. d Not included in Table 1 due to lack of dietary data for 1965. DIETARY FACTORS AND CORONARY HEART DISEASE TABLE 3 CORRELATION COEFFICIENTS BETWEEN THE CONSUMPTION OF VARIOUS DIETARY CONSTITUENTS a ToTotal oa SauteSimpl Complex Dietary constituents calorie fats fSaturaed Sucrose ' imple carbo- Proteins _intake s gars hydrate Total calorie intake Total fats Saturated fats Sucrose Simple sugars Complex carbohydrates Proteins 1.00 0.83 0.71 0.32 0.07 -0.59 0.27 0.83 1.00 0.90 0.60 0.30 -0.83 0.13 0.71 0.90 1.00 0.59 0.34 -0.83 0.18 0.32 0.60 0.59 1.00 0.59 -0.79 -0.36 0.07 0.30 0.34 0.59 1.00 -0.70 -0.42 -0.59 -0.83 -0.83 -0.79 -0.70 1.00 0.10 0.27 0.13 0.18 -0.36 -0.42 0.10 1.00 a Calculated from the data given in Table 1. and in death rates from AHD are rather poor. In other words, it seems that the death rates from AHD increased in the individual countries regardless of the changes that occurred in the calorie intake and in the consumption of the various dietary factors. DISCUSSION For many years investigators have sought an explanation for the differences between cardiovascu- lar mortality in developed and developing countries, and also for the increase that took place in practically all countries over the last decades. In is now widely held that many factors play a role in the etiology and pathogenesis of cardiovascu- lar disease; among them, diet has attracted the greatest interest and many authors believe that dietary factors play an important role in the develop- ment of atherosclerosis and of ischaemic heart disease. However, in spite of the great amount of epidemiological, clinical and experimental work that has been carried out by numerous investigators in attempting to define the role of diet in the develop- ment of the disease, it is still unclear to what extent diet can account for the observed phenomena and which dietary characteristics are responsible for the secular changes and the regional variations. It is therefore likely that other factors besides diet, those that are associated with living patterns, might influence the temporal and geographical differences in cardiovascular mortality. Using the findings of the present investigation as a guide-line, it will be useful to review briefly the literature and try to summarize the available know- ledge on the role of diet and other environmental factors in the development of AHD. It would, however, be beyond the scope of this article to provide detailed documentation; this can be found, for example, in the book published by Bajusz (1965) on the nutritional aspects of cardiovascular disease, in the review on dietary fat and coronary heart disease that was prepared by the National Heart Foundation of Australia (1967) and in the more recent publication on the geographical distribution of atherosclerosis in relation to diets and to other factors (Lab. Invest., 1968). Total calorie intake As shown in Table 1, calorie intake is positively associated with death rates from AHD on a geo- graphical basis. Albrink (1965) postulated that excessive calorie intake, rather than consumption of any particular kind of food, might be responsible for the higher incidence of AHD. According to this author, from an evolutionary point of view the high fat and protein consumption of modern times is indeed a harmless reversion to the primor- dial carnivorous habits of man; Albrink also con- sidered that the most important, recent change in human dietary habits is the greater calorie intake being enjoyed by certain populations. Proteins Most investigators consider that mortality from AHD is not related to the protein contribution to total calorie intake; however, some have demon- 107 108 R. MASIRONI TABLE 4 INCREASE IN AHD DEATH RATE AND IN CONSUMPTION OF VARIOUS DIETARY FACTORS IN RECENT YEARS Death rates Dietary factors (ratio 1965: 1955) (ratio 1962: 1947) Countries Males, Males, Both Total hSi pte55-64 years sexes, calorie Fat rated Sucrose sugars carbo- Protein yea a all ages i USA 1.04 1.04 1.09 0.95 1.08 0.93 0.97 1.00 0.84 0.96 Canada 1.05 1.10 1.06 0.95 1.08 0.94 1.12 0.81 0.86 1.00 Switzerland 1.05 1.06 0.92 1.04 1.28 1.18 1.15 0.95 0.73 0.93 Chile 1.10 1.37 1.47 1.02 1.36 1.14 - 1.31 0.92 1.02 Sweden 1.14 1.21 1.31 1.04 1.18 0.92 1.00 1.05 0.77 0.71 Jadpan 1.14 0.88 1.22 1.33 2.42 3.09 3.48 4.50 0.83 1.03 Australia 1.19 1.32 1.17 0.96 1.12 1.06 0.98 1.05 0.86 0.89 FiAnland 1.22 1.47 1.30 1.19 1.32 1.34 2.23 1.73 0.72 0.87 Italy 1.22 1.38 1.22 1.21 1.24 1.35 2.87 2.16 0.80 0.94 Belgiumn 1.22 1.30 0.97 1.14 1.37 1.04 1.07 1.08 0.73 0.94 United Kingdom 1.23 1.38 1.03 1.10 1.26 1.25 1.16 1.19 0.73 0.9 New Zealand 1.25 1.28 1.12 1.08 1.05 1.05 0.84 1.00 0.94 0.96 Federal ny 1.33 1.45 1.16 1.35 2.78 2.65 1.35 1.91 0.56 0.79 Denmark 1.38 1.36 1.36 1.08 1.14 1.12 1.32 1.36 0.73 0.80 Hungary 1.38 1.44 1.51 1.24 1.70 2.14 - 1.85 0.75 1.01 Uruguay 1.41 1.17 1.68 1.30 1.23 1.19 - 1.31 0.92 0.89 Netherlands 1.41 1.56 1.15 1.09 1.34 0.94 1.56 1.61 0.65 0.92 France 1.43 1.51 1.48 1.29 1.42 1.28 1.50 1.55 I 0.65 1.00 Greece 1.60 2.85 1.33 1.30 1.06 1.02 1.22 1.86 0.93 1.08 Norway 1.64 1.61 1.67 1.01 1.15 1.17 1.89 1.58 0.70 0.88 Poland 2.14 1.77 3.95 1.42 1.80 0.7 _ 3.6 0.68 1.0 Colombia 2.31 1.74 1.94 1.08 1.22 1.14 1.13 0.83 1.13 0.9 Yugoslavia 2.90 3.55 3.66 1.41 1.57 0.74 2.89 1.80 0.89 1.02 Mean 1.42 1.51 1.51 1.15 1.39 1.27 1.58 1.61 0.79 0.93 Correlation coefficients of death rate to dietary factors Males, 55-64 years of age 0.51 0.10 -0.25 0.29 0.18 0.26 0.26 Males, 45-54 years of age 0.48 -0.02 -0.31 0.24 0.07 0.20 0.30 Both sexes, all ages 0.57 0.18 -0.27 0.43 0.40 0.08 0.26 strated a positive relationship with consumption of factor in AHD and thrombosis. Most of the above animal protein (Yudkin, 1957; Jolliffe & Archer, authors, however, considered the relation between 1959; Annand, 1961, 1967; Scrimshaw & Guzman, protein and AHD to be a secondary association 1968). Annand (1967) considers that the consump- with no etiological significance. Experimental tion of heated animal proteins, particularly those studies on animals and in man, reported by Bajusz found in milk, may be an important etiological (1965) indicate that variations in dietary protein DIETARY FACTORS AND CORONARY HEART DISEASE intake have practically no effect on the development of atherosclerosis. Fats The finding by Keys (1953) that the fat content in the diet is positively correlated with the incidence of atherosclerosis and ischaemic heart disease gave rise to extensive speculation on the relationship existing between diet, the serum cholesterol level and atherosclerosis. Such a relationship does exist in experimental animals, where coronary athero- sclerosis and occlusions can indeed be produced by dietary manipulations (Friedman, 1960). Whether this relationship exists in man also is, however, still controversial. Many authors believe that excessive intake of dietary fats has an influence on the inci- dence of AHD (Higginson & Pepler, 1954; Bronte- Stewart, Keys & Brock, 1955; Keys et al., 1958; Paul et al., 1963; Epstein et al., 1965a, 1965b; Lopez et al., 1966; McGandy, Hegsted & Stare, 1967; Page & Stamler, 1968; Scrimshaw & Guzman, 1968), and strong evidence is available that the serum cholesterol level and incidence of coronary heart disease are directly correlated (Epstein, 1965; Stam- ler et al., 1960; Cornfield, 1962; Keys et al., 1963; Kannel et al., 1964; Morris et al., 1966; Scrimshaw & Guzman, 1968; Obeyesekere, 1968). Some investi- gations, however, showed no relation between the blood cholesterol level and dietary habits (Morris et al., 1963; Little et al., 1965). Several other authors, however, are critical of the role played by dietary fats in the development of the disease and they have revealed many instances in which the dietary fat intake, the blood cholesterol level, and the incidence of atherosclerosis or of coronary disease were not correlated: some examples may be found in the studies of Camara-Besa, Macapinlac & Albino (1959) on Filipino soldiers, of Karvonen et al. (1961) on Finnish lumberjacks and sedentary men, of Lapicirella et al. (1962) on certain Somali tribes, of Stuart et al. (1962) on the rural population of St. Kitts, of Shaper (1962, 1963) on the Samburu tribe of northern Kenya, of Mann et al. (1962, 1964) on Pygmies and Masai, of McDonough et al. (1965) on the high and low social classes of Evans County, Ga., USA, of Finegan et al. (1968) on cardiac patients and healthy subjects in Ireland, of Bassett & Schroff- ner (1968) on Chinese and Japanese men living in Hawaii, and of Brunner et al. (1969) on sedentary and non-sedentary workers in Israel. It should be pointed out, however, that several of the population groups mentioned above are habi- tually engaged in strenuous physical activity, and this might favourably influence the cardiovascular function. The allegedly protective role of physical activity will be discussed later in the text. Several authors (Sinclair, 1956; Kinsell et al., 1958; Keys et al., 1958; McGandy, Hegsted & Stare, 1967) maintain that the type, rather than the amount of fat consumed plays a role, and it has been demon- strated that saturated fats cause an increase in seum cholesterol level whereas unsaturated fats decrease it through mobilization of cholesterol (Kinsell & Michaels, 1955; Bronte-Stewart et al., 1956; Malm- ros & Wigand, 1957; Morris, 1960, 1961; Stamler, 1960). Christakis et al. (1966), Turpeinen ct al. (1968), Dayton et al. (1968) and others have indeed demonstrated that use of a cholesterol-lowering diet, in which some of the saturated fats have been replaced by unsaturated fats, brings about a decrease in the incidence of coronary heart disease in small, well-defined population groups. Other authors, however, do not share this view. According to the surveys carried out on Yemenites who immigrated into Israel (Cohen, Barly & Poz- nanski, 1961; Cohen, 1963) and on Polynesian popu- lations (Hunter, 1962; Prior et al., 1966) neither the amount nor the type of dietary fat seems to play a major role in the etiology of ischaemic heart disease. Antar, Ohlson & Hodges (1964) reported that the marked increase in coronary heart disease that has occurred in the USA over the past several decades was accompanied by only a slight (6 %) increase in fat consumption, due mostly to an increase in unsaturated fats. The ratio of polyunsaturated to saturated fatty acids, instead, increased markedly (29 %). These authors state that their data " do not fit the hypothesis that low ratios of polyunsaturated to saturated fatty acids in the diet are significantly related to high mortality rate from coronary heart disease ". Similar criticism against the hypothesis that saturated fatty acids have a primary role in the etiology of atherosclerosis had also been expressed by Annand (1961). According to Finegan et al. (1968), high calorie and saturated fat intakes may be risk factors at a population level, but they are probably risk factors of a low-grade type. The only conclusion that can be drawn from the examination of the literature and of the data pre- sented above is that although there is strong evidence that dietary fats, particularly the saturated ones, play an important role in the etiology of AHD, there is no proof that they are the only or the main culprit. 8 19 R. MASIRONI Carbohydrates The relation of carbohydrate intake to athero- sclerosis and ischaemic heart disease is also a source of much controversy. From an epidemiological point of view the following findings seem worthy of mention. Yudkin (1964) suggested that sugars may be the causal factor in the development of AHD and that the statistical association that several authors had demonstrated, for a number ofcountries, between fat intake and AHD may be only secondary to the association he found between the figures for national consumption of sugars and of fats. Osan- cova, Hejda & Zvolankova (1965) supported Yud- kin's hypothesis by demonstrating a good relation- ship between cardiovascular mortality and sugar consumption among the 10 administrative regions of Czechoslovakia, whereas fat consumption was poorly correlated. These authors, however, refrained from drawing any conclusions as to the role played by sugars in the development ofAHD. Cohen et al. (1961) and Cohen (1963) pointed out that the higher mortality from coronary heart disease in Yemenite settlers in Israel, as compared with their compatriots recently emigrated from the Yemen, may be related to their higher consumption of sugar. These authors postulated that sucrose may play a role in the devel- opment of ischaemic heart disease. One more piece of indirect evidence that sugars may be related to ischaemic heart disease is the finding by Antar, Ohlson & Hodges (1964), that the increased inci- dence of coronary heart disease in the USA over the past decades was accompanied by a marked increase in simple sugar consumption and by a greater decline in consumption of complex carbohydrates. These authors suggest that the changes that have occurred in the type of dietary carbohydrates may be a contributing factor to the increased incidence of coronary heart disease in the USA. Lopez et al. (1966) studied the relationship between food consumption and mortality from atherosclero- tic heart disease in Europe, and reported that " the highest correlation coefficient was found between the percentage of calories derived from simple carbo- hydrates and the death rate; this correlation has been positively maintained through the years covered by the study ". More recently, higher sugar consumption in a coronary group as compared to a non-coronary group was observed also by Paul et al. (1968), although the statistical analysis showed no significant association between heavy intake of sucrose and coronary disease. Additional, indirect, evidence also pointing to the possibility that carbohydrates may play a role in the development of atherosclerosis and myocardial infarction is the repeated finding that diabetes and decreased sugar tolerance are prominent risk factors of atherosclerotic heart disease. Clinical studies and epidemiological surveys confirmed that people affected by disorders of the carbohydrate metabolism are also more prone to myocardial infarction and to vascular lesions including athero- sclerosis (Stamler et al., 1960; Waddel & Field, 1960; Pell & D'Alonzo, 1961; Sowton, 1962; Yudkin & Roddy, 1964; Esptein et al., 1965a, 1965b; Ostrander et al., 1965; De Lange, Mey & Vivier, 1965; Kings- bury, 1966; Sievers, 1967; Epstein, 1967; Bassett & Schroffner, 1968). The possibility that excessive sugar consumption may be an etiological factor in the development of diabetes and of atherosclerosis was pointed out also by Ziegler (1966, 1967a, 1967b). These views, however, are challenged by Brunner & Altmann (1968), according to whom " no obligatory relationship exists between diabetes and athero- sclerosis ". As regards the relationship of sugars to cardio- vascular diseases it must be borne in mind that these nutrients have common metabolic pathways with fats. Disturbances in carbohydrate metabolism may be responsible for abnormal fat metabolism and may therefore act as a causative factor in the development of atherosclerosis and of coronary disease (De Lange, Mey & Vivier, 1965; Waddel & Field, 1960; Cohen, 1963; Falsetti et al., 1968). Minerals Another factor, also of nutritional importance, which it has recently been suggested may play a role in the etiology of cardiovascular diseases, is trace- element imbalance. Several authors (for references see Masironi, 1969) have repeatedly observed a statistical correlation between softness of drinking water and cardiovascular mortality rates; there is also limited evidence that various trace elements may influence cardievascular function and are involved in the pathogenesis of various cardio- vascular diseases. Man's mineral balance is changing as a function of technological progress; industrial pollution of air and water, the use of fertilizers and of food additives, food processing and canning, all alter the mineral balance because of accumulation of some trace ele- ments or depletion of others. Populations living in highly industrialized countries are more exposed to 110 DIETARY FACTORS AND CORONARY HEART DISEASE I1I artificial manipulations of the environment than are the populations of the developing countries, particu- larly in the past few decades in which the rate of technological progress increased markedly. It may be speculated that the differences in cardiovascular mortality rates that have been detected on an inter- national and on an intranational basis could be due, at least partly, to technologically-induced trace ele- ment imbalance. Although still very speculative, this new field of investigation is certainly worthy of consideration. Non-dietary factors From the data reported above and from the examination of other authors' findings the conclu- sion is that coronary heart disease cannot be attri- buted to any one single dietary factor. Although, in some studies, heavy consumption of fats and simple sugars has been found to be associated with high prevalence of atherosclerosis and ischaemic heart disease, in several other investigations no definite relationships to dietary habits (Paul et al., 1963; Papp, Padilla & Johnson, 1965; Finegan et al., 1968; Obeyesekere, 1968) were detected. Also in the context of a wide geographical investi- gation like the International Atherosclerosis Project it appears that in most population groups the sources of fats and of carbohydrates cannot be considered as the primary factors in determining severity of atherosclerotic lesions (Scrimshaw & Guzman, 1968). Such discrepancies may be explained, at least to a certain extent, by taking into consideration other interrelated factors that are always present, in addi- tion to the dietary ones, thus preventing a clear understanding of the relationship of diet to cardio- vascular diseases. These factors are: physical activity, psychological stresses and, perhaps, some of the concomitants of affluence. It was suggested by some authors (Lowenstein, 1964; Mann et al., 1964; Sievers, 1967) that the low death rates from ischaemic heart disease detected in population groups leading a rather primitive way of life could be attributed to the fact that these peoples are regularly engaged in strenuous physical activity, possess a good degree of physical fitness, and are free from the psychological stresses of a competitive society. It is beyond the scope of this article to discuss the effects of smoking on cardiovascular health. Although none of these factors individually may explain major differences in cardiovascular mortality rates, together they might play an impor- tant protective role. Physical activity has been reported to exert a favourable effect on the incidence of atherosclerosis and coronary heart disease (Morris et al., 1953a, 1953b; Karvonen et al., 1961; Mann et al., 1964; Brunner, 1966), possibly by lowering the blood lipid level (Lopez et al., 1966; Montoye et al., 1969). Mental stress, instead, allegedly provokes mobiliza- tion of nutritional deposits but these are not properly used: lipid levels in the blood increase (Bogdonoff et al.; 1 Steinberg & Shafrir, 1960) and this situation, if repeated or protracted for long periods of time, might precipitate the onset of AHD in people who, because of dietary, hereditary or other factors, are already prone to the disease (Charvat, Dell & Folkow, 1964). According to Russek (1967) psychological stress may enhance the atherogenicity and lethality of a high fat diet. The work of Groen et al. (1961), on Benedictine and Trappist monks, does indeed lend support to the hypothesis that a life free from social and psychological stress has, regardless of dietary patterns and habits, a beneficial effect on cardiovascular health. Observations on animals also showed the detri- mental effects of psycho-social stresses: hypercho- lesterolaemia, coronary atherosclerosis, myocardial necrosis, and myocardial infarction were produced in animals by subjecting them to emotional stresses (Uhley & Friedman, 1959; Groover et al., 1963; Raab, Chaplin & Bajusz, 1964; Ratcliffe & Snyder, 1967). It is true that the real importance of mental stress in the development ofAHD has never been definitely proved, and that people in primitive societies also experience considerable psychological stress but, as Jolliffe (1959) pointed out, the mental stress in modern civilization as opposed to the " stress of the jungle" is perhaps of a new type to which man has not yet learnt to adjust fully. Concerning prosperity, it seems to be associated with higher death rates from AHD on a geographical basis: if we take, for instance, the national per capita income and the national per capita energy consump- tion as indices of living standards and of technologi- cal progress, and relate them to the death rates from AHD, good correlations are found (see Table 1). Positive correlations of cardiovascular mortality to indices of national " development" such as income, number of television and radio sets, number of I Bogdonoff, M. D., Estes, E. H. & Harlan, W. (1959) Psychophysiologic studies of fat metabolism, a paper pre- sented at the 40th Annual Session of the American College of Physicians, Chicago, April 1959. R. MASIRONI telephones, etc., were also found by Yudkin (1957, 1964) and by Jolliffe & Archer (1959). Even on a small population basis it has been reported that prevalence of coronary heart disease increases with economic status (Bronte-Stewart, 1958; Morris, 1956; Servotham & Berry, 1968), although this association cannot be considered as well defined. Higher consumption of fats (particularly of the saturated type) and of refined sugars, and lower consumption of complex carbohydrates are usually associated with higher prosperity. Table 1 shows indeed that national per capita income correlates positively with intake of total and of saturated fats, and of sucrose, and negatively with consumption of complex carbohydrates. Therefore, it may well be that the relation between dietary factors and the incidence of AHD is secondary to other factors, which are in turn associated with living standards, and which perhaps play a more important role. In conclusion, the relationship of diet to athero- sclerosis and coronary heart disease is still a natter for further research; no single dietary factor, nor the blood cholesterol, may be held as the sole or even the main factor responsible for the development of the disease. AHD is currently regarded as a disease of multiple etiology in which dietary factors, obesity, elevated blood levels of cholesterol, tri- glycerides and other lipids, heredity, hypertension, diabetes mellitus, mental stress and physical inactiv- ity are considered to be important predisposing factors; when studying the possible influence of diet on cardiovascular health and disease, we should thus consider the way of life of the people. The way of life may cause the dietary factors to be harmful for some populations while they are harmless for others. Moreover, factors which are of a high-risk type on a population basis may not be so on an individual basis, and vice versa. ACKNOWLEDGEMENTS The author wishes to thank the following persons for having read and commented on the manuscript: Miss J. W. Marr, Nutritionist, Medical Research Council, SMU, London; Professor F. H. Epstein, Department of Epidemiology, University of Michigan, Ann Arbor, Mich., USA; Dr F. W. Lowenstein, Assistant Secretary, FAO/WHO/UNICEF Protein Advisory Group, United Nations, New York; Dr F. Jackson, Cardiovascular Department, Newcastle General Hospital, Newcastle- upon-Tyne, England; Professor N. S. Scrimshaw, Head, Department of Nutrition and Food Science, Massachu- setts Institute of Technology, Cambridge, Mass., USA; and Professor J. H. de Haas, Netherlands Institute for Preventive Medicine, Leiden, Netherlands. RESUME FACTEURS DIETETIQUES ET MALADIES DES CORONAIRES On a reuni pour 37 pays les informations (1965) relatives a la mortalite par arteriosclerose des coronaires et myocardite degenerative ainsi que les renseignements disponibles (1962) concernant la consommation per capita des graisses totales, des acides gras satures, du saccharose, des sucres simples et complexes, des proteines et l'apport de calories. Ces donnees ont ensuite fait l'objet d'une evaluation statistique afin de deceler une relation possible entre ces facteurs dietetiques et l'inci- dence des affections des coronaires. On releve une forte correlation positive entre la consom- mation de graisses totales et d'acides gras satures et les taux de mortalite par arteriosclerose des coronaires et myocardite degenerative. En ce qui regarde la consomma- tion de calories, de saccharoEe et de sucres simples, la correlation est moins nette et elle est quasiment nulle pour la consommation de prot6ines. La tres nette corr& lation negative existant entre la consommation d'hydrates de carbone complexes et la frequence des maladies des coronaires represente vraisemblablement une fausse relation etant donne qu'une correlation du meme type apparait aussi entre ce facteur et les autres facteurs diet& tiques, a l'exception des proteines. Des resultats simi- laires ont e obtenus par 1'analyse de donnees relatives a la decennie precedente. Les tendances exposees ci-dessus sont observees dans tous les groupes etudies, qui presentent diverses combi- naisons d'age et de sexe. On note cependant que les correlations sont parfois moins apparentes dans le sexe feminin. 11 n'existe par ailleurs qu'une tres faible correla- tion entre l'evolution des differents facteurs dietetiques et 1'augmentation croissante des taux de mortalite par arteriosclerose des coronaires et myocardite degenerative au cours des annees dans un pays donne. Les resultats de cette analyse sont examines et discutes a la lumiere des observations faites 'a l'occasion de nombreuses autres investigations. 11 semble que l'on puisse considerer comme etabli que 112 DIETARY FACTORS AND CORONARY HEART DISEASE 113 la consommation exageree de graisses, et surtout d'acides gras satur6s, favorise le developpement de I'arterioscle- rose des coronaires et de la myocardite degdndrative, mais une telle tendance ne se manifeste pas dans plusieurs groupes de population, et en particulier dans les couches sociales les plus defavorisees du point de vue economique et dans les pays en voie de developpement. I1 faut peut- etre en chercher la cause dans le fait que les individus appartenant a ces groupes ont une activite physique quo- tidienne relativement importante qui exercerait une sorte d'effet protecteur sur la fonction cardio-vasculaire. Les rapports entre le regime alimentaire, d'une part, et l'art6riosclerose des coronaires et la myocardite degene- rative, d'autre part, sont encore imprecis et l'apparition et le degr6 de gravite de ces affections ne peuvent etre attribues avec certitude a l'un ou l'autre facteur dietd- tique unique. 11 importe de tenir compte d'autres ele- ments, comme le manque d'activite physique, les tensions psycho-sociales et le bien-etre materiel, dont le r6le favorisant ne doit pas etre neglige. En ce qui regarde l'influence de la prosperite et du developpement techno- logique, on rel&ee, sur le plan national, de fortes correla- tions positives entre les taux de mortalite par arterioscle- rose des coronaires et myocardite degenerative et le revenu ou la consommation d'energie par tete d'habitant. L'influence eventuelle sur la fonction cardio-vasculaire des oligo-elements et le r6le du desequilibre de l'apport en sels mineraux resultant des modifications du milieu et des manipulations subies par les denrees alimentaires sont exposes a grands traits. REFERENCES Albrink, J. M. (1965) J. Amer. diet. Ass., 46, 26-29 Annand, J. C. (1961) J. Coll. gen. Practit., 4, 567- 596 Annand, J. C. (1967) J. Atheroscier. Res., 7, 797-801 Antar, M. A., Ohison, M. A. & Hodges, R. E. (1964) Amer. J. clin. Nutr., 14, 169-178 Bajusz, E. (1965) Nutritional aspects of cardiovascular diseases, London, Crosby Lockwood and Son, Ltd. Bassett, D. R. & Schroffner, W. (1968) Israel J. med. Sci., 4, 1087 Bronte-Stewart, B. (1958) Brit. med. Bull., 14, 243- 252 Bronte-Stewart, B., Antonis, A., Eales, L. & Brock, J. F. (1956) Lancet, 1, 521-526 Bronte-Stewart, B., Keys, A. & Brock, J. F. (1955) Lancet, 2, 1103-1107 Brunner, D. (1966) The influtence of physical activity on incidence and prognosis of ischaemic heart disease. In: Raab, W., ed., Prevention of ischaemic heart disease: principles and practice, Springfield, Ill., Charles C. Thomas Brunner, D. & Altman, S. (1968) Vascular diseases in diabetes. In: Proceedings of the Ninth International Congress of Life Assurance Medicine, Tel A viv, 1967, Basel, Karger Brunner, D., Altman, S., Loebl, K. & Schwartz, S. (1970) J. chron. Dis., (in press) Camara-Besa, S., Macapinlac, M. P., and Albino, A. M. (1959) Amer. J. clin. Nutr., 7, 702-710 Charvat, J., Dell, P. & Folkow, B. (1964) Cardiologia (Basel), 44, 124-141 Christakis, G. et al. (1966) Amer. J. publ. Haelth, 56, 299-314 Cohen, A. M., Bavly, S. & Poznanski, R. (1961) Lancet, 2, 1399-1401 Cohen, A. M. (1963) Amer. Heart J., 65, 291-293 Cornfield, J. (1962) Fed. Proc., 21, 58-61 Dayton, S. et al. (1968) Lancet, 2, 1060-1062 De Lange, D. J., Mey, H. S. & Vivier, F. S. (1965) S. Afr. med. J., 39, 119 Den Hartog, C., Buzina, K., Fidanza, F., Keys, A., & Roine, P. (1968) Dietary studies and epidemiology of heart diseases, The Hague, Stichting tot Wetenschap- pelijke Voorlichting op Voedingsbied Epstein, F. H. (1965) J. chron. Dis., 18, 735-774 Epstein, F. H. (1967) Circulation, 36, 609-619 Epstein, F. H., Francis, T., Jr, Hayner, N. H., Johnson, B. C., Kjelsberg, M. O., Napier, J. A., Ostrander, L. D,. Payne, N. W. & Dodge, H. J. (1965a) Amer. J. Epidem. 81, 307-322 Epstein, F. H., Ostrander, L. D., Johnson, B. C., Payne, M. W., Hayner, N. S., Keller, J. B. & Frances, T., Jr (1965b) Ann. intern. Med., 62, 1170-1187 Falsetti, H. L., Schnatz, J. D., Greene, D. G. & Bunnell, I. L. (1968) Circulation, 37, 184-191 Finegan, A., Hickey, N., Maurer, B. & Mulcahy, R. (1968). Amer. J. clin. Nutr., 21, 143-148 Food and Agriculture Organization (1949) Food balance sheets, 1947-48, Rome Food and Agriculture Organization (1953) Food coinposi- tion tables for initernational use, Rome Food and Agriculture Organization (1966) Food balance sheets, 1960-62, Rome Friedman, G. J. (1960) Modern nutrition in health and disease, Philadelphia, Wohl and Goodhart, p. 774 Groen, J. J., Tijong, B. K., Koster, M., Verdonck, G., Pierloot, R. & Willenbrandt, A. F. (1961) Ned. T. Geneesk., 105, 222-233 Groover, M. E., Jr, Seljeskog, E. L., Haglin, J. J. & Hitchcock, C. R. (1963) J. Angiology, 14, 409 Haas, J. H. de (1964) Changing mortality patterns and cardiovascular diseases, Haarlem, De Erven F. Bohn N.V. Higginson, J. & Pepler, W. J. (1954) J. clin. Iduest., 33, 1366-1371 114 R. MASIRONI Hunter, J. D. (1962) Fed. Proc., 21, 36-43 Jolliffe, N. (1959) Circulation, 20, 109-127 Jolliffe, N. (1961) Metabolism, 10, 497-513 Jolliffe, N. & Archer, M. (1959) J. chron. Dis., 9, 636-651 Kannel, W. B., Dawber, T. R., Friedman, G. D., Glen- non, W. E. & McNamara, P. M. (1964) Ann. intern. Med., 61, 888-899 Karvonen, M. J., Rantaharju, P. M., Orma, E., Punsar, S. & Takkunen, J. (1961) J. occup. Med., 3, 49-53 Keys, A. (1953) J. Mt Sinai Hospital, 20, 118-139 Keys, A. (1956) J. Chron. Dis., 4, 364-380 Keys, A., Kimura, N., Kusukawa, A., Bronte-Stewart, B., Larsen, N. & Keys, M. H. (1958) Ann. intern. Med., 48, 83-94 Keys, A. et al. (1963) Circulation, 28, 381-395 Kingsbury, K. J. (1966) Lancet, 2, 1374-1379 Kinsell, L. W. & Michaels, G. D. (1955) Amer. J. clin. Nutr., 3, 247-253 Kinsell, L. W. et al. (1958) Lancet, 1, 334-339 Lab. Invest., 1968, 18, 463-653 Lapiccirella, V., Lapiccirella, R., Abboni, F. & Liotta, S. (1962) Bull. Wld Hlth Org., 27, 681-697 Lew, E. A. (1957) J. chron. Dis., 6, 192-209 Little, J. A., Shanoff, H. M., Csima, A., Redmund, S. E. & Yano, R. (1965) Lancet, 1, 933-935 Lopez, A., Krehl, W. A., Hodges, R. E. & Good, E. I. (1966) Amer. J. clin. Nutr., 19, 361-369 Lowenstein, F. W. (1964) Amer. J. clin. Nutr., 15, 175-186 McDonough, J. R., Hames, C. G., Stulb, S. C. & Garri- son, G. E. (1965) J. chron. Dis., 18, 443-468 McGandy, R. B., Hegsted, D. M. & Stare, F. J. (1967) New Engl. J. Med., 277, 186-192, 242-247 Malmros, H. & Wigand, G. (1957) Lancet, 2, 1-7 Mann, G. V., Roels, 0. A., Price, D. L. & Merrill, J. M. (1962) J. chron. Dis., 15, 341-371 Mann, G. V., Shaffer, R. D., Anderson, R. S. & Sand- stead, H. H. (1964) J. Atheroscler. Res., 4, 289-312 Masironi, R. (1969) Bull. WId HIth Org., 40, 305-312 Montoye, H. J. et al. (1959) Amer. J. clin. Nutr., 7, 139-145 Morris, J. N. (1956) Incidence of coronary disease in population groups in England. In: Keys, A. & White, P. D., ed., World Trends in Cardiology, Vol. I. Cardio- vascular Epidemiology, New York, Paul B. Hoeber, Inc. p. 42-49 Morris, J. N. (1960) Mod. Conc. cardiov. Dis., 29, 625-632, 30, 633-638 Morris, J. N. (1961) Mod. Conc. cardiov. Dis., 30, 633-638 Morris, J. N., Heady, J., Raffle, P., Roberts, C. & Parks, J. (1953a) Lancet, 2, 1053-1057 Morris, J. N., Heady, J., Raffle, P., Roberts, C. & Parks, J. (1953b) Lancet, 2, 1111-1120 Morris, J. N., Kagan, A., Pattison, D. C., Gardner, M. J. & Raffle, P. (1966) Lancet, 2, 553-559 Morris, J. N., Marr, J. W., Heady, J. A., Mills, G. L. & Pilkington, T. R. E. (1963) Brit. med. J., 1, 571-576 National Heart Foundation of Australia (1967) Med. J. Aust., 54, 309-322 Obeyesekere, I. (1968) Israel J. med. Sci., 4, 1110 Oliver, M F. & Stuart-Harris, C. H. (1965) Brit. med. J., 2, 1203-1208 Osancova, K., Hejda, S. & Zvolankova, K. (1965) Lancet, 1, 494 Ostrander, L. D., Francis, T., Hayner, N. S. Kjelsberg, M. 0. & Epstein, F. H. (1965) Ann. intern. Med., 62, 1188-1198 Page, H. I. & Stamler, J. (1968) Mod. Conc. cardiov. Dis., 37, 119-128 Papp, 0. A., Padilla, L. & Johnson, A. L. (1965) Lancet, 2, 259-261 Paul, O., McMillan, A., McKean, H. & Park, H. (1968) Lancet, 2, 1049-1051 Paul, 0. et al. (1963) Circulation, 28, 20-31 Pell, S. & D'Alonzo, C. A. (1961) J. Amer. med. Ass., 175, 463-470 Prior, I. A. M. et al. (1966) The health of two groups of Cook Island Maori, Wellington, Medical Research Council of New Zealand Raab, W., Chaplin, J. P. & Bajusz, E. (1964) Proc. Soc. exp. Biol. (N. Y.), 116, 665 Ratcliffe, H. L. & Snyder, R. L. (1967) Brit. J. exp. Path., 48, 357-365 Reader, R. & Wynn, A. (1966) Med. J. Aust., 53, 740-743 Russek, H. I. (1967) Dis. Chest, 52, 1-9 Scrimshaw, N. S. & Guzman, M. A. (1968) Lab. Invest., 18, 623-628 Servotham, S. G. & Berry, J. N. (1968) Circulation, 37, 939-953 Shaper, A. G. (1962) Amer. Heart J., 63, 437-442 Shaper, A. G. (1963) Amer. J. clini. Nutr., 13, 135-146 Sievers, M. L. (1967) Ann. intern. Med., 67, 800-807 Sinclair, H. M. (1956) Lancet, 1, 381-383 Sowton, E. (1962) Brit. med. J., 1, 84-86 Stamler, J. et al. (1960) J. chron. Dis., 11, 405-419 Steinberg, D. & Shafrir, E. (1960) Med. News, 6, 1 Stuart, K. L. et al. (1962) Brit. med. J., 2, 1283-1288 Tejada, C., Strong, J. P., Montenegro, M. R., Restrepo, C. & Solberg, L. A. (1968) Lab. Invest., 18, 509-526 Turpeinen, 0. et al. (1968) Amer. J. cliti. Nutr., 21, 255-276 Uhley, H. N. & Friedman, M. (1959) Amer. J. Phiysiol., 197, 396 United Nations (1966) Yearbook of national account statistics, 1966, New York United Nations (1967) Statistical yearbook, 1967, New York Waddel, W. R. & Field, R. A. (1960) Metabolism, 9, 800 World Health Organization (1958) World health statistics annual, 1955, Geneva World Health Organization (1968) World health statistics annual, 1965, Geneva Yudkin, J. (1957) Lancet, 2, 155-162 Yudkin, J. (1964) Lancet, 2, 4-5 Yudkin, J. & Roddy, J. (1964) Lancet, 2, 6-8 Ziegler, E. (1966) Schweiz. med. Wschr., 96, 1345-1349 Ziegler, E. (1967a) Helv. paediat. Acta, 22, 360-365 Ziegler, E. (1967b) Mitt. natur. Ges. Winterthur, 32, 67-72